Hudson Robert J, Thomson Ian R, Jassal Rajive, Peterson David J, Brown Aaron D, Freedman Jeffrey I
Department of Anesthesia, St. Boniface General Hospital, University of Manitoba, Winnipeg, Canada.
Anesthesiology. 2003 Oct;99(4):847-54. doi: 10.1097/00000542-200310000-00016.
Although fentanyl has been widely used in cardiac anesthesia, no complete pharmacokinetic model that has assessed the effect of cardiopulmonary bypass (CPB) and that has adequate predictive accuracy has been defined. The aims of this investigation were to determine whether CPB had a clinically significant impact on fentanyl pharmacokinetics and to determine the simplest model that accurately predicts fentanyl concentrations during cardiac surgery using CPB.
Population pharmacokinetic modeling was applied to concentration-versus-time data from 61 patients undergoing coronary artery bypass grafting using CPB. Predictive ability of models was assessed by calculating bias (prediction error), accuracy (absolute prediction error), and measured:predicted concentration ratios versus time. The predictive ability of a simple three-compartment model with no covariates was initially compared to models with premedication (lorazepam vs. clonidine), sex, or weight as covariates. This simple model was then compared to 18 CPB-adjusted models that allowed for step changes in pharmacokinetic parameters at the start and/or end of CPB. The predictive ability of the final model was assessed prospectively in a second group of 29 patients.
None of the covariate (premedication, sex, weight) models nor any of the CPB-adjusted models significantly improved prediction error or absolute prediction error, compared to the simple three-compartment model. Thus, the simple three-compartment model was selected as the final model. Prospective assessment of this model yielded a median prediction error of +3.8%, with a median absolute prediction error of 15.8%. The model parameters were as follows: V1, 14.4 l; V2, 36.4 l; V3, 169 l; Cl1, 0.82 l. min-1; Cl2, 2.31 l x min-1; Cl3, 1.35 l x min-1.
Compared to other factors that cause pharmacokinetic variability, the effect of CPB on fentanyl kinetics is clinically insignificant. A simple three-compartment model accurately predicts fentanyl concentrations throughout surgery using CPB.
